CARL: a running recognition algorithm for free-living accelerometer data

Wearable accelerometers hold great promise for physical activity epidemiology and sports biomechanists. However, identifying and extracting data from specific physical activities, such as running, remains challenging. Objective: To develop and validate an algorithm to identify bouts of running in raw, free-living accelerometer data from devices worn at the wrist or torso (waist, hip, chest). Approach: The CARL (continuous amplitude running logistic) classifier identifies acceleration data with amplitude and frequency characteristics consistent with running. The CARL classifier was trained on data from 31 adults wearing accelerometers on the waist and wrist, then validated on free-living data from 30 new, unseen subjects plus 166 subjects from previously-published datasets using different devices, wear locations, and sample frequencies. Main Results: On free-living data, the CARL classifier achieved mean accuracy (F1 score) of 0.984 (95% confidence interval 0.962-0.996) for data from the waist and 0.994 (95% CI 0.991-0.996) for data from the wrist. In previously-published datasets, the CARL classifier identified running with mean accuracy (F1 score) of 0.861 (95% CI 0.836-0.884) for data from the chest, 0.911 (95% CI 0.884-0.937) for data from the hip, 0.916 (95% CI 0.877-0.948) for data from the waist, and 0.870 (95% CI 0.834-0.903) for data from the wrist. Misclassification primarily occurred during activities with similar torso acceleration profiles to running, such as rope jumping and elliptical machine use. Significance: The CARL classifier can accurately identify bouts of running as short as three seconds in free-living accelerometry data. An open-source implementation of the CARL classifier is available at <<GITHUBURL>>.

Free-Living Gait Cadence Measured by Wearable Accelerometers for Assessing Fall Risk

Accelerometers are widespread in research applications, yet whether they are superior to structured clinic-based assessments is unknown. Using negative binomial regression, we compared traditional in-clinic measures of mobility (6-minute gait cadence, speed, and distance, and 4-meter gait speed) with free-living gait cadence from wrist accelerometers (Actigraph GT9X) in predicting fall rates in 432 older adults (age 77.29±5.46 yrs, 59.1% men, 80.2% White) participating in the Study to Understand Fall Reduction and Vitamin D in You (STURDY). Accelerometry-based gait cadence was estimated with the Adaptive Empirical Pattern Transformation algorithm. Across all participants, every 10 steps/min higher cadence was associated with a 13.2% lower fall rate (p=0.036). Mobility measures were not related to falls (p&gt;0.05). Among higher-functioning participants (cadence ≥100 steps/min), every 10 steps/min higher free-living cadence (p=0.01) was associated with a 27.7% lower fall rate. Data collected from accelerometers may provide a more sensitive indicator of fall risk than in-clinic tests.

Novel Applications of Accelerometry Data for Health Outcomes in Older Adults: Thinking Beyond MVPA

Physical activity is a well-established predictor of health and longevity. Wearable accelerometers produce high-frequency, time series data that capture multiple aspects of daily physical activity across the spectrum of intensity. Historically, the majority of accelerometry-based physical activity research has employed summary threshold metrics such as moderate-to-vigorous physical activity, or “MVPA.” Although these measures are important for understanding compliance with physical activity guidelines, they underutilize the potential of this data. To advance the science of physical activity in older adults, more sensitive, clinically translatable measures are needed. This symposium will examine the associations between novel measures of accelerometry-derived physical activity and various aging-related health outcomes. Dr. Wanigatunga will discuss the association of physical activity volume and fragmentation with the frailty phenotype in the Study to Understand Vitamin D and Fall Reduction in You (STURDY). Dr. Cai will present evidence on the association of physical activity quantities and patterns with measures of visual impairment in the Baltimore Longitudinal Study of Aging. Ms. Qiao will present a novel accelerometry-derived measure of performance fatigability in the Developmental Epidemiologic Cohort Study. Finally, Dr. Urbanek will discuss the role of accelerometry-derived free-living gait cadence in defining fall risk in STURDY. Collectively, these presentations highlight critical associations between objective measures of physical activity and health outcomes in older adults and illuminate the need for thinking beyond MVPA to improve prevention and intervention efforts.

A Novel Accelerometry-Based Metric to Improve Estimation of Whole-Body Mechanical Load

While the Player Load is a widely-used parameter for physical demand quantification using wearable accelerometers, its calculation is subjected to potential errors related to rotational changes of the reference frame. The aims of this study were (i) to assess the concurrent validity of accelerometry-based Player Load against force plates; (ii) to validate a novel metric, the Accel’Rate overcoming this theoretical issue. Twenty-one recreational athlete males instrumented with two triaxial accelerometers positioned at the upper and lower back performed running-based locomotor movements at low and high intensity over six in-series force plates. We examined the validity of the Player Load and the Accel’Rate by using force plates. Standard error of the estimate was small to moderate for all tested conditions (Player Load: 0.45 to 0.87; Accel’Rate: 0.25 to 0.95). Accel’Rate displayed trivial to small mean biases (−1.0 to 6.1 a.u.) while the Player Load displayed systematic very large to extremely large mean biases (17.1 to 226.0 a.u.). These findings demonstrate a better concurrent validity of the Accel’Rate compared to the Player Load. This metric could be used to improve the estimation of whole-body mechanical load, easily accessible in sport training and competition settings.

Walking stability in patients with benign paroxysmal positional vertigo: an objective assessment using wearable accelerometers and machine learning

Background Benign paroxysmal positional vertigo (BPPV) is one of the most common peripheral vestibular disorders leading to balance difficulties and increased fall risks. This study aims to investigate the walking stability of BPPV patients in clinical settings and propose a machine-learning-based classification method for determining the severity of gait disturbances of BPPV. Methods Twenty-seven BPPV outpatients and twenty-seven healthy subjects completed level walking trials at self-preferred speed in clinical settings while wearing two accelerometers on the head and lower trunk, respectively. Temporo-spatial variables and six walking stability related variables [root mean square (RMS), harmonic ratio (HR), gait variability, step/stride regularity, and gait symmetry] derived from the acceleration signals were analyzed. A support vector machine model (SVM) based on the gait variables of BPPV patients were developed to differentiate patients from healthy controls and classify the handicapping effects of dizziness imposed by BPPV. Results The results showed that BPPV patients employed a conservative gait and significantly reduced walking stability compared to the healthy controls. Significant different mediolateral HR at the lower trunk and anteroposterior step regularity at the head were found in BPPV patients among mild, moderate, and severe DHI (dizziness handicap inventory) subgroups. SVM classification achieved promising accuracies with area under the curve (AUC) of 0.78, 0.83, 0.85 and 0.96 respectively for differentiating patients from healthy controls and classifying the three stages of DHI subgroups. Study results suggest that the proposed gait analysis that is based on the coupling of wearable accelerometers and machine learning provides an objective approach for assessing gait disturbances and handicapping effects of dizziness imposed by BPPV.

Validation of Walking Speed Estimation from Trunk Mounted Accelerometers for a Range of Walking Speeds

Walking speed is a strong indicator of the health status of older people and patients. Using algorithms, the walking speed can be estimated from wearable accelerometers, which enables minimally obtrusive (longitudinal) monitoring. We evaluated the performance of two algorithms, the inverted pendulum (IP) algorithm, and a novel adaptation correcting for lateral step movement, which aimed to improve accuracy during slow walking. To evaluate robustness, we gathered data from different groups (healthy adults, elderly, and elderly patients) of volunteers (n = 159) walking under various conditions (over ground, treadmill, using walking aids) at a broad range of speeds (0.11–1.93 m/s). Both of the algorithms showed good agreement with the reference values and similar root-mean-square errors (RMSEs) for walking speeds ≥0.5 m/s, which ranged from 0.09–0.16 m/s for the different positions, in line with the results from others. However, for slower walking, RMSEs were significantly better for the new method (0.06–0.09 m/s versus 0.15–0.19 m/s). Pearson correlation improved for speeds <0.5 m/s (from 0.67–0.72 to 0.73–0.82) as well as higher speeds (0.87–0.97 to 0.90–0.98) with the new method. Overall, we found that IP(-based) walking speed estimation proved to be applicable for a variety of wearing positions, conditions and speeds, indicating its potential value for health assessment applications.

Diurnal Physical Activity Patterns across Ages in a Large UK Based Cohort: The UK Biobank Study

The ability of individuals to engage in physical activity is a critical component of overall health and quality of life. However, there is a natural decline in physical activity associated with the aging process. Establishing normative trends of physical activity in aging populations is essential to developing public health guidelines and informing clinical perspectives regarding individuals’ levels of physical activity. Beyond overall quantity of physical activity, patterns regarding the timing of activity provide additional insights into latent health status. Wearable accelerometers, paired with statistical methods from functional data analysis, provide the means to estimate diurnal patterns in physical activity. To date, these methods have been only applied to study aging trends in populations based in the United States. Here, we apply curve registration and functional regression to 24 h activity profiles for 88,793 men (N = 39,255) and women (N = 49,538) ages 42–78 from the UK Biobank accelerometer study to understand how physical activity patterns vary across ages and by gender. Our analysis finds that daily patterns in both the volume of physical activity and probability of being active change with age, and that there are marked gender differences in these trends. This work represents the largest-ever population analyzed using tools of this kind, and suggest that aging trends in physical activity are reproducible in different populations across countries.

Quantifying the Varying Predictive Value of Physical Activity Measures Obtained from Wearable Accelerometers on All-Cause Mortality over Short to Medium Time Horizons in NHANES 2003–2006

Physical activity measures derived from wearable accelerometers have been shown to be highly predictive of all-cause mortality. Prediction models based on traditional risk factors and accelerometry-derived physical activity measures are developed for five time horizons. The data set contains 2978 study participants between 50 and 85 years old with an average of 13.08 years of follow-up in the NHANES 2003–2004 and 2005–2006. Univariate and multivariate logistic regression models were fit separately for five datasets for one- to five-year all-cause mortality as outcome (number of events 46, 94, 155, 218, and 297, respectively). In univariate models the total activity count (TAC) was ranked first in all five horizons (AUC between 0.831 and 0.774) while the active to sedentary transition probability (ASTP) was ranked second for one- to four-year mortality models and fourth for the five-year all-cause mortality model (AUC between 0.825 and 0.735). In multivariate models age and ASTP were significant in all one- to five-year all-cause mortality prediction models. Physical activity measures are consistently among the top predictors, even after adjusting for demographic and lifestyle variables. Physical activity measures are strong stand-alone predictors and substantially improve the prediction performance of models based on traditional risk factors.